cps-attack
Audited by Socket on Sep 19, 2026
4 alerts found:
Securityx4SUSPICIOUS/HIGH-RISK skill. Its capabilities are internally consistent with its stated purpose, but that purpose is offensive exploitation of cyber-physical systems, including credential extraction, PLC manipulation, and safety-system bypass with real-world impact. No strong malware indicators or hidden exfiltration endpoints are present, and the flagged curl examples are documentation, but the overall skill is dangerous because it operationalizes offensive ICS attacks for an AI agent.
The fragment is a highly actionable ICS attack and compromise playbook. It contains explicit procedures for credential theft, authentication bypass, PLC/HMI/SIS manipulation, malicious firmware deployment, OT-to-IT pivoting, and sensitive-data extraction. It is not obfuscated and does not by itself prove executable malware or package-install-time behavior, but its operational security risk is very high if distributed as part of a software package or used without authorization.
The artifact is an offensive ICS/OT red-team playbook, not executable package code and not evidence of npm supply-chain malware. It contains highly actionable instructions for PLC manipulation, covert persistence, safety-data spoofing, historian extraction, and OT-to-IT pivoting. Although authorization and safety constraints are stated, the examples could cause physical disruption or conceal unsafe conditions if misused. It should be restricted to explicitly authorized, isolated testing and reviewed for removal or redaction of production attack steps.
The shown portion is an offensive ICS/OT penetration-testing procedure document, not ordinary dependency implementation code. It contains numerous high-impact commands capable of unauthorized access, sensitive-data extraction, OT-to-IT pivoting, persistence, alarm suppression, and SIS manipulation. Its stated assessment context and remediation guidance are consistent with authorized testing, so there is insufficient evidence that the fragment itself contains malware. Nevertheless, it presents substantial operational and security risk if distributed or executed without explicit authorization and isolated lab controls.